Design Of Linear Feature Matching Method Based On Node Similarity

With the continuous development and widespread application of GIS technology,the demand for spatial data is increasing day by day.In order to meet people's different needs,the departments of surveying and mapping have collected and produced a large number of various spatial data,Which are influenced by the factors such as collecting time,collection means,collection personnel and scale,and there are significant differences in data accuracy,spatial location,geometry,and attribute information,which makes "multi-source heterogeneous spatial data management and utilization" become a major problem,with issues such as integration and fusion,sharing,updating,and poor quality of spatial data being the most prominent.As one of the key technologies to solve the above problems,the homonym entity matching technology has become a hot topic at home and abroad,and has great research significance and application value.As the main type of spatial data,the linear feature are rich in information,easy to extract and describe,and have become the most active research direction in the study of homonymous entity matching.This paper takes the linear feature as the research object,and comprehensively summarizes the development and research of linear feature at home and abroad.The deficiencies of existing research are pointed out,and a new method of linear feature matching based on node similarity is proposed.This method is based on the principle of spatial similarity,taking the points of linear feature as the main research object,selecting the topological features and geometric features to match the homonymous entity,using the similarity degree of evaluation indexes to carry out layer constraints,and finally the optimal matching is obtained by calculating the spatial similarity values of the homonymous nodes.In order to achieve the above methods,the following aspects are studied in this paper.(1)The research background and significance of the paper are expounded,the status of the research on the matching of linear feature at home and abroad is analyzed in detail,and the shortcomings of the existing research are summarized,and the research content and technical route of this paper are formulated.(2)Based on the principle of spatial similarity of homonymous entity,the technique of homonymous entity matching is introduced in detail,the common methods and strategies for linear feature matching are summarized,and their advantages and disadvantages are analyzed.At the same time,complex network theory is introduced into linear feature matching.A linear feature matching method which based on node similarity is proposed.(3)A similarity matching model of linear feature is designed.For the problems of linear feature with the same topology but different structure,and linear feature have strong stability at the intersection,the similarity index which include the degree of linear feature node,the angle and the distance,and the similarity matching model which bases on topological relationship and geometric features is constructed.The similarity matching model is used to filter the candidate matching set by three constraints: node similarity,direction similarity,and position similarity.The angle and distance are weighted and normalized to calculate the comprehensive value of spatial similarity,the value is used to get the homonymous node,and complete the matching of linear feature.(4)Based on this model,a matching algorithm of linear feature is designed and is optimized.Among them,the use of bidirectional matching strategy effectively solves multiple complex situations in linear feature matching and improves the accuracy of the matching algorithm.In order to quickly acquire the candidate matching set,the method based on buffer is used,and improves the efficiency of the matching algorithm.(5)A linear feature matching prototype system is designed and implemented.Based on the Windows environment,using an object-oriented programming method and related component technology,a matching experiment system with C/S architecture of linear feature is built.The road network data in same scale and multi-scale are used to experiment,the feasibility of the matching method is verified.The experiment results show that the matching method of line feature based on node similarity is feasible.In addition,this method has been successfully applied to the actual project,achieving the data correction of a large amount of data which is in 1:250000 scale,the method has greatly improved the efficiency of data production and has a strong practical value.